Adjuvant irradiation is currently the standardtreatment for many malignant neoplasms. Gyne- cological cancer is the most common pelvic malig-nancy receiving radiotherapy, followed by gastroin-

J Soc Colon Rectal Surgeon (Taiwan) September 2011

Case Analysis

Overview of Radiation-Related Colovesical

or Rectovesical Fistula: Experience from

18 Patients at the Chang Gung Memorial

Hospital in Kaohsiung

Kai-Lung Tsai

Chien-Chang Lu

Hong-Hwa Chen

Shung-Eing Lin

Division of Colon and Rectal Surgery,

Department of Surgery, Chang-Gung

Memorial Hospital � Kaohsiung Medical

Center, Chang-Gung University College of

Medicine, Kaohsiung, Taiwan

Key Words

Radiation;

Colovesical fistula;

Rectovesical fistula

Background. Adjuvant irradiation is currently the standard treatment formany malignant neoplasms; however, late complication due to radiation isa clinically common problem. This report presents our experience withlong-term follow up for radiation-related colovesical or rectovesical fis-tula at the Kaohsiung Chang-Gung Memorial Hospital.

Methods. Eighteen patients (16 female, 2 male), from 1989 to 2003, withradiation-related colovesical or rectovesical fistula were retrospectivelyanalyzed. The clinical course, pathologic findings, associated treatment,

and outcomes were assessed.

Results. Patients were diagnosed with rectal cancer (n = 2, 11.1%), cer-vical cancer (n = 12, 66.7%), bladder cancer (n = 2, 11.1%), endometrialcancer (n = 1, 5.5%), and simultaneous bladder and cervical cancers (n =1, 5.5%). The diagnostic tools included barium enema, cystography, ab-dominal computed tomography, colonoscopy and fistulography. And thesensitivity rate is 55.6%, 44.4%, 42.9%, 33.3% and 0% respectively. Sixpatients received diverting colostomy including 1 concomitant ileal con-duit, 10 patients received Hartmann’s procedure including 2 concomitantsmall bowel segmental resections, and 1 patient received repair of fistulathrough laparotomy. Almost all patients received symptom relief afteroperation, but 3 patients developed recurrent fistulas.

Conclusion. The barium enema is the most accurate diagnostic tool. Ad-ditionally, a combination of gastrointestinal and urological surgery isusually needed because of the nature of fistulas. One-stage operationsshould be limited to younger patients in good nutritional states and with-out severe inflammation, radiation injury, intestinal obstruction, or othermajor medical problem such as advanced malignancy.[J Soc Colon Rectal Surgeon (Taiwan) 2011;22:79-85]

Received: December 1, 2010. Accepted: May 27, 2011.

Correspondence to: Dr. Shung-Eing Lin, Chien-Chang Lu, Division of Colon and Rectal Surgery, Department of Surgery,

Chang-Gung Memorial Hospital, No. 123, Ta-Pei Road, Niao-Sung Hsiang, Kaohsiung 833, Taiwan. Tel: +886-7-731-7123 ext. 8008;

Fax: +886-7-735-4309; E-mail: doctorlu@adm.cgmh.org.tw

79

testinal malignancies, male genital cancer, and carci-

nomas of the bladder.1,2

Late-onset radiotherapy-induced complications

occur in 51% of patients, including bowel and urinary

bladder complications in 41% and 21%, respectively.3

Chronic gastrointestinal complications include cram-

ping abdominal pain, diarrhea, or cachexia, or may

present acutely as bowel obstruction or fistula.4 Ra-

diation-related mortality in small bowel disease is

32%, while that for colonic disease is 4%.5 Urological

complications include cystitis, hematuria, ureteral ste-

nosis, and fistula formation.6,7

Long-term follow up data for radiation-induced

colovesical or rectovesical fistula is important be-

cause radiation may lead to late-onset complications.

Detecting and identifying the long-term complica-

tions associated with radiation is also crucial. This re-

port presents our experience with long-term follow up

for radiation-related colovesical or rectovesical fistu-

las at the Kaohsiung Chang-Gung Memorial Hospital.

Patients and Methods

A total of 18 patients were retrospectively en-

rolled from 90 patients with enterovesical fistulas. Pa-

tients were excluded because of benign disease (n =

33), cancer invasion (n = 35), and enterovesical fis-

tulas (n = 4). Medical records for the remaining 18

patients (16 female, 2 male) with radiation-related

colovesical or rectovesical fistula between 1989 and

2003 were retrospectively analyzed. The last follow

up was in June 2008. The mean age was 61.5 years old

(range 36-75 years). Fifteen patients (83.3%) received

operations and adjuvant radiotherapy, 1 underwent

surgery and concurrent radiochemotherapy, and 2 re-

ceived radiotherapy only. The mean radiation dose

was 4648 cGy (range 500-9900 cGy). The patients’

demographic data is shown in Table 1.

Results

Among the 18 patients (16 women and 2 men)

with radiation-related colovesical or rectovesical fis-

tulas, the mean age was 61.5 (10.9) years (range 36-75

years) with female predominance (88.9%). Patients

were diagnosed with rectal cancer (n = 2, 11.1%), cer-

vical cancer (n = 12, 66.7%), bladder cancer (n = 2,

11.1%), endometrial cancer (n = 1, 5.5%), and simul-

taneous bladder and cervical cancers (n = 1, 5.5%).

All patients received radiotherapy with or without

surgical intervention (15 patients underwent surgery

and 1 received concurrent chemotherapy). Seven pa-

tients showed recurrent cancer or distant metastasis at

the time of the diagnosis of fistula. The mean duration

from diagnosis of previous disease to the diagnosis of

colovesical or rectovesical fistula was 8.8 years (range

3-18 years). Six patients had systemic diseases, in-

cluding 3 with diabetes mellitus, 1 with hypertension,

and 2 with both diseases.

The clinical symptoms are summarized in Table 2.

We found fecaluria in 6 patients (33.3%), fever or

chillness in 8 patients (44.4%), dysuria in 5 patients

(27.8%), hematuria in 3 patients (16.7%), urine from

the rectum in 2 patients (11.1%), urine leakage from

the abdominal wound in 2 patients (11.1%), stool

from the vagina in 1 patient (5.6%), bloody stool in 1

patient (5.6%) and abdominal pain in 2 patients

(11.1%). Pneumaturia was not recorded in any of the

patients.

Urine analysis was done in only 16 patients; of

80 Kai-Lung Tsai, et al. J Soc Colon Rectal Surgeon (Taiwan) September 2011

Table 1. Demographic data of 18 patients with colovesical or

rectovesical fistula

Mean age (years) 61.5 (10.9) n = 18

Male 2 11.1%

Female 160 88.9%

Predisposing disease

Cervical ca 120 66.7%

Bladder ca 2 11.1%

Rectal ca 2 11.1%

Endometrial ca 1 05.6%

Cervical and bladder ca 1 05.6%

Underlying disease

DM 5 83.3%

HTN 3 50%

Treatment

Medication 1 05.6%

Hartmann’s procedure 100 55.6%

Segmental resection of small bowel 2 11.1%

Ileal conduit 1 05.6%

Diverting stoma 6 33.3%

Repair of fistula 1 05.6%

these, 13 patients (81.2%) had WBC > 5/mm3 and 3

had WBC < 5/mm3. Urine cultures were taken for 16

patients, 10 of which proved to have gram-negative

bacterial infections. The most common pathogen

was Escherichia coli (80%), followed by Pseudomo-

nas aeruginosa (30%), Proteus sp. (10%), and

Klebsiella pneumoniae (10%). Six patients (37.5%)

showed renal insufficiency. The mean white count

was 10,606/ mm3 (range 4600-19,800/mm3). Ab-

dominal CT was performed in 7 patients, of which 3

were positive for fistula. Seven patients underwent

cytoscopy with 3 positive findings of fistula. Cys-

tography was performed in 9 patients, of which 4

positive findings of fistula were seen. Barium enema

was done in 9 patients with 5 positive findings of fis-

tula. Colonoscopy was done in 3 patients with 1 posi-

tive findings of fistula. Fistulography was done in 3

patients without positive findings of fistula. Positive

prediction rates of different diagnostic tools are sum-

marized in Table 3.

All patients received surgical intervention, and

different procedures were chosen. Six patients re-

ceived diverting colostomy including 1 concomitant

ileal conduit, 10 patients received Hartmann’s proce-

dure including 2 concomitant small bowel segmental

resections, and 1 patient received repair of fistula via

the abdomen. The orifice of the fistula was found at

the intestinal end in 13 patients, including 6 at the

rectum, and 7 at the colon respectively. Postopera-

tively, 1 patient had short bowel syndrome. Sixteen

patients have stoma present after the operation, and

no reconstruction was done in any of these cases, no

matter with or without urological surgery (ileal con-

duit). Three mortalities were noted at 3 months of

follow up in this group. Almost all patients received

symptom relief after operation, but 3 patients deve-

loped recurrent fistulas.

Discussion

Colovesical fistula is a rare disease and is one of

the late-onset complications of pelvic radiotherapy.

The etiologies of colovesical fistula are known to be

cancer invasion, irradiation, diverticulitis, inflamma-

tory bowel disease, or others. In one study, the under-

lying pathology included diverticular disease (62%),

carcinoma (27%), and inflammatory bowel disease

(6%).8 In Hayne et al.’s study, 9 more than 75% of pa-

tients receiving pelvic radiotherapy experienced acute

anorectal symptoms and up to 20% suffered from

late-phase radiation proctitis. About 5% developed

other chronic complications such as fistulas, stricture,

and disabling fecal incontinence.

Solkar et al. retrospectively reviewed 50 cases of

colovesical fistula with etiologies of diverticular dis-

ease, Crohn’s disease, and pelvic malignancy.10 The

most common clinical presentations were fecaluria,

pneumaturia, or both. In our study, pneumaturia was

never recorded. However, this may be due to the dif-

ficulty in recognizing the symptom by patients.

Predisposing factors of radiation-induced compli-

cations include radiation dosage, type and mode of

delivery, pelvic inflammatory disease, hypertension,

diabetes, arteriosclerotic heart disease, and age greater

than 50 years.11 In our study, the mean age of the pa-

tients was 61.5 years. Some of our patients had sys-

temic diseases such as diabetes mellitus and hyper-

tension. The mean total dose of radiation to treat pel-

vic cancer was 4648 cGy. Hong et al. concluded that

low-pelvic RT alone (range 40-60 Gy [median 50 Gy])

significantly reduces the complications in node-nega-

Vol. 22, No. 3 Radiation-Related Colovesical or Rectovesical Fistula 81

Table 3. Diagnostic examination made in 18 patients

Diagnostic tool

No. of patients

with positive

finding of fistula

No. of patients

receiving

examination

Sensitivity

(%)

Barium enema 5 9 55.6%

Cystography 4 9 44.4%

Abdominal CT 3 7 42.9%

Colonoscopy 1 3 33.3%

Fistulography 0 3 0

Table 2. Clinical symptoms of 18 patients with colovesical or

rectovesical fistula

Symptoms

Fever or chills 8 44.4%

Fecaluria 6 33%

Dysuria 5 27.8%

Hematuria 3 16.7%

Abdominal pain 2 11.1%

Urine from anus 2 11.1%

Urine from abdominal wound 2 11.1%

Stool from vagina 1 05.6%

Bloody stool 1 05.6%

tive, high-risk, Stage I-IIA cervical cancer patients.12

The lag time from radiotherapy to the develop-

ment of fistulas was previously unclear. In our study,

the mean time for induction of colovesical fistulas

was found to be 8.8 years after radiotherapy, which

may suggest long-term follow up to confirm this late-

onset complication of radiotherapy.

Traditionally, cystoscopy and barium enema exam-

ination are most commonly used modality to confirm

the presence of colovesical fistulas.13 Cystoscopy

would demonstrate abnormalities in 90% of cases and

the fistulas might be declared in 33% of cases (Fig. 1).

Biopsies should also be taken to exclude a urological

malignancy as the cause of fistula. Barium enema

shows a fistula in only a third of the cases and can help

determine underlying colorectal disease. More re-

cently, computed tomography (CT) has been proposed

to be an effective diagnostic investigative tool; it has a

diagnostic accuracy of more than 90%. Jarrett and

Vaughan described that abdominal CT helps confirm

the diagnosis and provides important anatomical infor-

mation by revealing the intraluminal and extraluminal

pathological status.14 In a review study by Sarr et al.,

the most common abnormal findings in CT included

intravesical gas (90%), focal bladder wall thickening

(90%), and an extraluminal mass (75%) (Fig. 2).15 Al-

though the scan did not always show the fistulous tract,

it accurately predicted the correct location in all cases.

82 Kai-Lung Tsai, et al. J Soc Colon Rectal Surgeon (Taiwan) September 2011

Fig. 1. A. Cytoscopy images showing stool content in the urinary bladder. B. Colon shown by contrast through the urinarybladder.

Fig. 2. KUB and abdominal CT images showing intravesical air. Intravesical air and thickening of urinary bladder wall arethe typical CT findings for enterovesical fistula.

Daniels et al. recommended cystoscopy and urine

cytology for fecal material as first-line investigations

in all patients with a suspected enterovesical fistula.16

CT scanning and barium enema should not be first-

line investigations, but may be performed subsequ-

ently to help determine etiology and to plan surgery.

Other methods had also been used to confirm the

presence of colovesical fistulas. Volkmer et al. used

transrectal 3D ultrasound with contrast media to de-

tect fistulas.17 Wrong surgical choice may leave re-

sidual disease for enterovesicle fistulas and therefore,

definite confirmation of the location of fistulas before

operation is very important.

The purpose of surgery is to reduce gastroin-

testinal and urological symptoms. The principle of

surgical procedure is to separate the gastrointestinal

and genitourinary tracts or can even be primary clo-

sure of the fistula. The procedure may consist of one

stage, which was first described in 1972, or multiple

stages, depending to the condition of the patients.18-20

The one-stage method is currently the management of

choice in patients with no sepsis, obstruction, or organ

impairment and in those with a normal nutritional sta-

tus, especially in patients with an inflammatory cause

of the fistula, diverticular, or granulamatous bowel

disease. Staged repairs may be more judicious in pa-

tients with advanced malignancy, radiation changes,

severe inflammation, large pelvic abscesses, or inade-

quate bowel preparation.21,22

The most common selected procedure to treat co-

lovesical or rectovesial fistulas is sigmoid colectomy

followed by Hartmann’s procedure and anterior resec-

tion.21-23 However, this course of action should be

limited to younger patients in good nutritional states

and without severe inflammation, radiation injury, in-

testinal obstruction, and other major medical pro-

blems such as advanced malignancy.23 In our practice,

less invasive procedures were favored because of the

fragility of radiation-injured tissue. Segreti et al. con-

cluded that there were no statistically significant dif-

ferences between patients treated with loop or end co-

lostomy with regard to early morbidity or survival.24

In our study, 10 patients received end colostomy and 6

patients received loop colostomy. There is not enough

data to support which procedure is superior, but there

is evidence that loop colostomy leaves residual dis-

ease. There were 3 patients with recurrent fistulas

after treatment, 1 of which underwent medical treat-

ment only, and 1 who received repair of fistula via the

abdomen.

In summary, colovesical or rectovesical fistula is a

late-onset complication following radiotherapy for

pelvic malignancies due to either tumor recurrence or

the effect of radiation. Several kinds of diagnostic

tools may help confirm the presence and nature of the

fistula, and therefore careful choosing of the appro-

priate tool is important. Diverting stoma is usually the

preferred procedure. Additionally, a combination of

gastrointestinal and urological surgery is usually needed

because of the nature of fistulas.

References

1. Martel P, Deslandes M, Dugue L, Sezeur A, Gallot D,

Malafosse M. Radiation injuries of the small intestine. Ann

Chir 1996;50:312-7.

2. Dean RJ, Lytton B. Urologic complications of pelvic irradia-

tion. J Urol 1978;119:64-7.

3. Jereczek-Fossa B, Jassem J, Nowak R, Badzio A. Late com-

plications after postoperative radiotherapy in endometrial

cancer: analysis of 317 consecutive cases with application of

linear-quadratic model. Int J Radiat Oncol Biol Phys 1998;

41:329-38.

4. Mann WJ. Surgical management of radiation enteropathy.

Surg Clin North Am 1991;71:977-90.

5. Kwitko AO, Pieterse AS, Hecker R, Rowland R, Wigg DR.

Chronic radiation injury to the intestine: a clinico-patho-

logical study. Aust N Z J Med 1982;12:272-7.

6. Holler U, Hoecht S, Wudel E, Hinkelbein W. Osteoradio-

necrosis after radiotherapy for gynecologic tumors. Strah-

lenther Onkol 2001;177:291-5.

7. Gellrich J, Hakenberg OW, Oehlschlager S, Wirth MP. Mani-

festation, latency and management of late urological com-

plications after curative radiotherapy for cervical carcinoma.

Onkologie 2003;26:334-40.

8. Driver CP, Anderson DN, Findlay K, Keenan RA, Davidson

AI. Vesico-colic fistulae in the Grampian region: presenta-

tion, assessment, management and outcome. J R Coll Surg

Edinb 1997;42:182-5.

9. Hayne D, Vaizey CJ, Boulos PB. Anorectal injury following

pelvic radiotherapy. Br J Surg 2001;88:1037-48.

10. Solkar MH, Forshaw MJ, Sankararajah D, Stewart M, Parker

MC. Colovesical fistula � is a surgical approach always

justified? Colorectal Dis 2005;7:467-71.

11. Pesce F, Righetti R, Rubilotta E, Artibani W. Vesico-crural

and vesicorectal fistulas 13 years after radiotherapy for pro-

Vol. 22, No. 3 Radiation-Related Colovesical or Rectovesical Fistula 83

state cancer. J Urol 2002;168:2118-9.

12. Hong JH, Tsai CS, Lai CH, Chang TC, Wang CC, Lee SP,

Tseng CJ, Hsueh S. Postoperative low-pelvic irradiation for

stage I-IIA cervical cancer patients with risk factors other

than pelvic lymph node metastasis. Int J Radiat Oncol Biol

Phys 2002;53:1284-90.

13. Sandison A, Jones PA. Urine culture in the diagnosis of

colovesical fistula. BMJ 1993;307:1588.

14. Jarrett TW, Vaughan ED Jr. Accuracy of computerized to-

mography in the diagnosis of colovesical fistula secondary to

diverticular disease. J Urol 1995;153:44-6.

15. Sarr MG, Fishman EK, Goldman SM, Siegelman SS, Cameron

JL. Enterovesical fistula. Surg Gynecol Obstet 1987;164:

41-8.

16. Daniels IR, Bekdash B, Scott HJ, Marks CG, Donaldson DR.

Diagnostic lessons learnt from a series of enterovesical

fistulae. Colorectal Dis 2002;4:459-62.

17. Volkmer BG, Nesslauer T, Kufer R, Loffler M, Maier S,

Gottfried HW. Diagnosis of vesico-intestinal fistulas by con-

trast medium enhanced 3-D ultrasound. Ultraschall Med

2001;22:81-6.

18. Carpenter WS, Allaben RD, Kambouris AA. One-stage

resections for colovesical fistulas. J Urol 1972;108:265-7.

19. Shackley DC, Brew CJ, Bryden AA, Anderson ID, Carlson

GL, Scott NA, Clarke NW. The staged management of com-

plex entero-urinary fistulae. BJU Int 2000;86:624-9.

20. McBeath RB, Schiff M Jr, Allen V, Bottaccini MR, Miller JI,

Ehreth JT. A 12-year experience with enterovesical fistulas.

Urology 1994;44:661-5.

21. Moss RL, Ryan JA Jr. Management of enterovesical fistulas.

Am J Surg 1990;159:514-7.

22. Mileski WJ, Joehl RJ, Rege RV, Nahrwold DL. One-stage re-

section and anastomosis in the management of colovesical

fistula. Am J Surg 1987;153:75-9.

23. Hsieh JH, Chen WS, Jiang JK, Lin TC, Lin JK, Hsu H.

Enterovesical fistula: 10 years experience. Zhonghua Yi Xue

Za Zhi (Taipei) 1997;59:283-8.

24. Segreti EM, Levenback C, Morris M, Lucas KR, Gershenson

DM, Burke TW. A comparison of end and loop colostomy for

fecal diversion in gynecologic patients with colonic fistulas.

Gynecol Oncol 1996;60:49-53.

84 Kai-Lung Tsai, et al. J Soc Colon Rectal Surgeon (Taiwan) September 2011

蔡鎧隆等 J Soc Colon Rectal Surgeon (Taiwan) 2011;22:79-85 85

病例分析

放射治療相關之大腸/直腸-膀胱廔管：高雄長庚 18例的經驗回顧

蔡鎧隆 盧建璋 陳鴻華 林尚潁

高雄長庚紀念醫院 大腸直腸肛門外科

背景 目前放射治療已經廣泛應用在多種癌症的治療，且放射治療相關的遲發性併發症也是臨床上常見的問題。本篇文章針對本院對放射治療相關之大腸/直腸-膀胱廔管的經驗與長期追蹤結果提供分享。

方法 本研究蒐集自 1989 至 2003 年間，在本院有紀錄之放射治療相關的大腸/直腸-膀胱廔管，吾人等分析其臨床上的病史、病理診斷、治療、以及治療的成果。

結果 在這 18例病患中，分別因以下原因接受放射治療；直腸癌 (n = 2, 11.1%)、子宮頸癌 (n = 12, 66.7%)、膀胱癌 (n = 2, 11.1%)、子宮內膜癌 (n = 1, 5.5%)，另有一位同時患有膀胱癌及子宮頸癌 (n = 1, 5.5%)。常用的診斷工具包括鋇劑攝影、膀胱攝影、腹部電腦斷層、大腸鏡和廔管攝影，檢查的敏感度分別為 55.6%，44.4%，42.9%，33.3% 和0%。共有 6 名患者接受分流性大腸造廔，其中有一位同時接受人工膀胱手術；有 10 名患者接受 Hartmann’s 手術，其中有兩位同時接受部分小腸切除，有一位接受經腹部廔管修補手術。幾乎所有患者在接受手術治療之後都得道症狀緩解，但有 3 例發生廔管復發。

結論 鋇劑攝影是正確率最高的診斷工具。考量疾病的成因，腸道及泌尿道的合併手術經常是必要的。一次性手術僅侷限於少部分病患適用；年輕、營養狀況佳、沒有進行性

的發炎、沒有放射線相關損傷、沒有腸阻塞或其他主要內科疾病，如侵犯性的癌症。

關鍵詞 放射線、結腸-膀胱廔管、直腸-膀胱廔管。